INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
Web-Based Boarding House Management Information System with  
Dashboard Analytics and Multiple Linear Regression for Rental  
Income Prediction  
Riky Fauzan*, Syarif Hidayatulloh  
Information Systems Study Program, Faculty of Information Technology, Universitas Adhirajasa  
Reswara Sanjaya, Bandung, Indonesia  
DOI: https://doi.org/10.51583/IJLTEMAS.2026.150600293  
Received: 18 July 2026; Accepted: 23 July 2026; Published: 04 August 2026  
ABSTRACT  
Small boarding-house businesses commonly rely on receipts and annual notebooks to record tenants, rental  
periods, payments, unit availability, and income. Such practices can support daily operations, but they make  
historical retrieval, operational monitoring, and decision support increasingly difficult as the number of records  
grows. This study develops a web-based boarding house management information system that integrates  
operational data management, dashboard analytics, and rental-income prediction for Surapati Boarding House  
in Bandung, Indonesia. The application was developed with Laravel 12, PHP, and MySQL, while the predictive  
model was trained offline in Python using scikit-learn. Because the original tenant and payment records contain  
private information, a simulated historical dataset was constructed from the property’s operational  
characteristics. The dataset contains 54 monthly observations from January 2022 to June 2026. The independent  
variables were occupied Type A units, occupied Type B units, and payment arrears, while monthly rental income  
was the dependent variable. After cleaning currency-formatted values and removing invalid rows, the data were  
divided into 43 training observations and 11 testing observations. The resulting multiple linear regression model  
was integrated into the Laravel analytics page through its estimated intercept and coefficients. The system  
provides authentication, unit and tenant master data, rental transactions, payment records, operational  
summaries, due-date reminders, and an analytics interface that compares actual and predicted income. Model  
evaluation produced an R-squared value of 95.85%, a mean absolute error of IDR 250,715.83, a root mean  
squared error of IDR 303,605.01, and a mean absolute percentage error of 1.39%. These results indicate close  
agreement between predicted and simulated actual income. However, the findings are limited to the simulated  
case-study dataset and require validation using anonymized real operational data before broader deployment.  
Keywords: boarding house management system, dashboard analytics, multiple linear regression, rental income  
prediction, web application  
INTRODUCTION  
Boarding-house management involves tenant registration, unit availability monitoring, rental-period  
administration, payment recording, and income reporting. When these activities are handled through receipts  
and handwritten notebooks, information can be scattered across documents and retrieving older records becomes  
time-consuming. Web-based information systems have therefore been applied to boarding-house administration  
to centralize tenant, unit, transaction, and payment data [2], [3], [5], [10], [13].  
Surapati Boarding House is a small rental business in Bandung that manages 12 rental units: five Type A units  
with two bedrooms and seven Type B units with one bedroom. The owner previously recorded tenant identities  
and payments using receipts and annual notebooks. Although the manual process remained usable, it became  
difficult to retrieve records from several years earlier and to obtain a rapid overview of unit occupancy, due  
dates, arrears, and income trends.  
Page 3944  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
A dashboard can summarize operational information through key indicators and visualizations, allowing  
managers to understand conditions without reviewing every transaction individually. Prior research has shown  
that well-designed dashboards can improve access to operational insights and support decision-making [1], [6],  
[14]. In addition, historical operational data can be used to estimate future or scenario-based income through  
regression models [4], [11].  
Previous boarding-house systems have largely focused on digital administration, tenant records, payments,  
complaints, and unit information [2], [3], [5], [10], [13]. The contribution of this study is the integration of  
management functions, dashboard analytics, and a multiple linear regression module in one web application.  
The study aims to (1) develop an integrated web-based management system, (2) provide concise operational  
analytics, and (3) implement and evaluate an income-prediction model using the available case-study variables.  
Related Work  
Setiawan et al. developed a web-based boarding-house information system to support rental data management  
[13]. Basir and Saifudin similarly reported a web system for boarding-house information and administration [3],  
while Hasti and Zaelani implemented website-based tenant, complaint, and payment management [5]. These  
studies demonstrate the operational value of digitization but do not combine administrative functions with  
predictive income analytics.  
Nugraheny et al. applied multiple linear regression in a web-based minimum-income prediction application [11],  
and Fadianty used multiple linear regression to predict sales turnover from historical data [4]. These studies  
support the use of regression for numeric business prediction. The present study adapts the approach to a small  
boarding-house context and integrates model outputs into a Laravel dashboard.  
METHODOLOGY  
Research design and system development  
The study used a case-study and prototype-oriented system-development approach. Requirements were collected  
through observation of the existing administrative process and interviews with the boarding-house owner. The  
prototype model was selected because screen layouts and workflows could be evaluated and refined against the  
owner’s operational needs [9].  
The application was implemented with Laravel 12 using the Model-View-Controller architecture, PHP for  
server-side processing, MySQL for persistent storage, and Laragon as the local development environment. The  
system contains authentication, rental-unit master data, tenant master data, rental transactions, payments, an  
operational dashboard, and an analytics report. Laravel’s framework structure was used to separate business  
logic, views, and database interaction [8].  
Figure 1: System architecture and offline prediction-model integration.  
Page 3945  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
Dataset and variables  
Real tenant and payment records were not directly used because they contain personally identifiable information.  
Instead, a simulated historical dataset was generated from the boarding house’s operational characteristics,  
including the fixed unit inventory, rental rates, occupancy changes, contract renewals, departures, new tenants,  
payment schedules, and arrears. The dataset contains 54 monthly observations covering January 2022 through  
June 2026.  
Type A represents occupied two-bedroom rental units and Type B represents occupied one-bedroom rental units.  
The target variable is monthly rental income. Payment inflow was retained as operational information but was  
not selected as a predictor because tenants may pay three, six, or twelve months in advance, causing payment  
inflow to be zero in some months even when units remain occupied.  
Table 1: Variables used in the multiple linear regression model.  
Variable  
Description  
Role  
X1  
X2  
X3  
Y
Number of occupied Type A units  
Number of occupied Type B units  
Payment arrears (IDR)  
Independent variable  
Independent variable  
Independent variable  
Dependent variable  
Monthly rental income (IDR)  
Data preprocessing and model training  
The spreadsheet contained currency values represented as text, such as “Rp14.750.000”. Preprocessing removed  
empty header rows and fully empty records, standardized the column names, stripped currency symbols and  
thousands separators, converted numeric attributes to floating-point values, and removed rows with incomplete  
model variables. All 54 valid monthly observations were retained.  
The data were divided into 43 training observations and 11 testing observations using an 80:20 holdout split  
implemented with scikit-learn’s train_test_split function (random_state = 42). A LinearRegression estimator was  
fitted using ordinary least squares, which minimizes the sum of squared residuals [7], [12]. Model training was  
performed offline in Python; the estimated intercept and coefficients were then inserted into the Laravel analytics  
controller for real-time scenario calculations.  
Ŷ = a + b₁X₁ + b₂X₂ + b₃X₃  
Evaluation metrics  
Prediction performance was evaluated on the testing set using the coefficient of determination (R²), mean  
absolute error (MAE), root mean squared error (RMSE), and mean absolute percentage error (MAPE). R²  
measures the proportion of variation explained by the model, MAE reports the average absolute error in rupiah,  
RMSE gives greater weight to large errors, and MAPE expresses average absolute error as a percentage.  
RESULTS  
System implementation  
The implemented application centralizes data that were previously recorded in separate paper documents. The  
main dashboard displays monthly income, occupied and available units, units under maintenance, recent income  
trends, and upcoming rental due dates. Master-data pages allow the administrator to manage units and tenants,  
while operational pages record rental agreements and payments.  
Page 3946  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
The interface was designed for a single administrator role representing the owner or manager. Functional  
verification was performed for login, data creation and editing, rental transactions, payment recording, dashboard  
summaries, and prediction calculations.  
Figure 2: Main operational dashboard of the boarding-house management system.  
Regression model  
The trained model produced the parameters shown in Table 2. The negative arrears coefficient indicates that,  
while Type A and Type B occupancy are held constant, an increase of IDR 1,000,000 in arrears is associated  
with an estimated decrease of approximately IDR 63,490 in recognized monthly income.  
Table 2: Estimated multiple linear regression parameters.  
Parameter  
Estimated value  
2,306,841.69  
2,309,775.01  
1,098,862.53  
−0.063490  
Intercept (a)  
Type A coefficient (b1)  
Type B coefficient (b2)  
Arrears coefficient (b3)  
Ŷ = 2,306,841.69 + 2,309,775.01X₁ + 1,098,862.53X₂ − 0.063490X₃  
Prediction results  
Table 3 presents the complete 11-observation testing set in chronological order. The absolute difference between  
actual and predicted income ranges from IDR 51,108.10 to IDR 600,029.37.  
Table 3: Actual and predicted income on the testing set (IDR).  
Period  
Actual income  
16,750,000.00  
16,750,000.00  
18,500,000.00  
18,500,000.00  
Predicted income  
17,040,254.37  
17,040,254.37  
18,139,116.90  
18,139,116.90  
Absolute error  
290,254.37  
290,254.37  
360,883.10  
360,883.10  
Apr-2022  
Jun-2022  
Jan-2023  
Feb-2023  
Page 3947  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
Jun-2023  
Aug-2023  
Sep-2024  
Sep-2025  
Jan-2026  
Feb-2026  
May-2026  
18,750,000.00  
16,750,000.00  
18,500,000.00  
20,500,000.00  
20,500,000.00  
20,500,000.00  
20,500,000.00  
19,350,029.37  
17,040,254.37  
18,139,116.90  
20,448,891.90  
20,448,891.90  
20,448,891.90  
20,448,891.90  
600,029.37  
290,254.37  
360,883.10  
51,108.10  
51,108.10  
51,108.10  
51,108.10  
Figure 3: Comparison of actual and predicted income for the testing observations.  
Model evaluation  
Table 4: Model evaluation results.  
Metric  
R²  
Result  
95.85%  
MAE  
RMSE  
MAPE  
IDR 250,715.83  
IDR 303,605.01  
1.39%  
The R² score indicates that the model explains 95.85% of the variation in the testing data. The MAPE of 1.39%  
indicates a low relative error for the simulated dataset, while MAE and RMSE show that the average monetary  
deviation is small relative to monthly income values of approximately IDR 14.75–20.50 million.  
Analytics-dashboard integration  
The analytics page presents R², MAPE, the number of historical observations, a prediction simulator, and a  
comparison chart. Users enter occupied Type A units, occupied Type B units, and arrears; the page calculates  
income using the stored regression parameters. For example, five occupied Type A units, six occupied Type B  
units, and no arrears produce an estimated income of approximately IDR 20.45 million.  
The prediction module uses offline training rather than calling Python as a live service. This reduces deployment  
complexity for a small local application, although retraining requires rerunning the Python script and updating  
the stored parameters.  
Page 3948  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
Figure 4: Integrated prediction and analytics page showing model metrics and scenario-based income  
estimation.  
DISCUSSION  
The developed system extends earlier boarding-house applications by combining administrative management,  
operational visualization, and numeric income prediction in a single application. Where prior studies primarily  
addressed tenant data, payments, and unit information [2], [3], [5], [10], [13], the present implementation adds  
an analytics page that exposes model performance and allows scenario-based estimates.  
The dashboard condenses operational data into cards, reminders, and charts. This design is consistent with  
research emphasizing that dashboards should provide actionable information rather than reproduce raw  
transactional records [6], [14]. For the case-study owner, the integrated view reduces the need to search annual  
notebooks to determine occupancy, upcoming due dates, or income trends.  
The regression results demonstrate close numerical agreement on the simulated testing set. However, the high  
R² and low MAPE must be interpreted cautiously. The dataset was constructed from a fixed inventory and known  
rental-price structure; occupied Type A and Type B units are therefore direct drivers of monthly income.  
Consequently, the model primarily captures the internal logic of the simulation rather than unexpected market  
behavior. The findings establish that the implementation pipeline works, but they do not prove equivalent  
predictive accuracy on real-world data.  
The inclusion of arrears provides an operational adjustment, but additional predictors such as maintenance  
downtime, price changes, seasonality, contract renewal probability, and actual anonymized cash-flow timing  
could improve realism. A chronological holdout or rolling time-series evaluation should also be considered in  
future work to better represent deployment in which earlier periods predict later periods.  
The offline integration strategy is practical for a small boarding-house system because it avoids maintaining a  
separate Python web service. Its limitation is that the model does not update automatically when new records  
are added. A future version could store model metadata in MySQL, trigger scheduled retraining, and preserve  
model versions for auditability. Future implementation may further improve adaptability by applying automated  
retraining schedules and storing updated model artifacts within the database environment.  
The current evaluation should be interpreted as an initial feasibility assessment because the model validation  
was performed using a simulated dataset derived from a single boarding-house case study. Future validation  
Page 3949  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
using anonymized real-world rental transactions and additional external variables, such as seasonal demand  
changes and maintenance conditions, is required to assess broader generalization performance.  
CONCLUSION  
This study developed a web-based boarding-house management information system for Surapati Boarding  
House using Laravel 12 and MySQL. The application integrates unit and tenant data, rental transactions,  
payments, operational summaries, due-date reminders, and analytics in one interface.  
A multiple linear regression model was trained in Python using 54 monthly simulated observations. Occupied  
Type A units, occupied Type B units, and payment arrears were used to estimate monthly rental income. On the  
11-observation testing set, the model achieved R² of 95.85%, MAE of IDR 250,715.83, RMSE of IDR  
303,605.01, and MAPE of 1.39%. The estimated parameters were successfully embedded in the Laravel  
analytics page for interactive scenario calculations.  
The system demonstrates a feasible approach to combining operational information management and lightweight  
predictive analytics for a small rental business. Because the evaluation used simulated data derived from the  
case-study structure, future research should validate the model with anonymized real transactions, use time-  
aware validation, and compare multiple prediction methods before operational decisions rely on its estimates.  
Data Availability  
The dataset used in this study is a simulated monthly dataset based on the operational characteristics of the case-  
study boarding house. It does not contain identifiable tenant information. The processed dataset and Python  
analysis script may be obtained from the corresponding author upon reasonable request.  
Conflict of Interest  
The authors declare no conflict of interest.  
REFERENCES  
1. Ahad, A., & Assegaff, S. (2022). Rancang bangun sistem informasi eksekutif dashboard  
monitoring produksi. 7(2), 300–314.  
2. Arigie, M., Fadillah, R., & Wahyuni, S. N. (2024). Pengembangan sistem informasi pengelolaan  
kos berbasis website dengan menggunakan framework CodeIgniter pada Wisma Mutiara Selaras.  
1(2), 37–45.  
3. Basir, A., & Saifudin, A. (2022). Rancang bangun sistem informasi rumah kos berbasis web dengan  
metode  
terapan.  
Jurnal  
Teknologi  
Sistem  
Informasi,  
5(2),  
139–144.  
https://doi.org/10.32493/jtsi.v5i2.20691  
4. Fadianty, R. (2024). Penerapan data mining dengan algoritma regresi linear berganda untuk  
memprediksi omset penjualan minyak goreng. Building of Informatics, Technology and Science,  
6(2). https://doi.org/10.47065/bits.v6i2.5951  
5. Hasti, N., & Zaelani, M. I. (2023). Sistem informasi pengelolaan Rumah Kost Budi Sari berbasis  
website. 2(2), 26–32.  
6. Hjelle, S., Mikalef, P., Altwaijry, N., & Parida, V. (2024). Organizational decision making and  
analytics: An experimental study on dashboard visualizations. Information & Management, 61(6),  
104011. https://doi.org/10.1016/j.im.2024.104011  
7. James, G., Witten, D., Hastie, T., Tibshirani, R., & Taylor, J. (2023). An introduction to statistical  
learning, with applications in Python. Springer. https://doi.org/10.1007/978-3-031-38747-0  
8. Laravel. (2025). Laravel documentation. https://laravel.com/docs  
Page 3950  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue VI, June 2026  
9. Maryani, M., et al. (2022). Comparison of the system development life cycle and prototype model  
for software engineering. International Journal of Emerging Technology and Advanced  
Engineering, 12(4), 155–162. https://doi.org/10.46338/ijetae0422  
10. Nizar, C. (2021). Rancang bangun sistem informasi sewa rumah kost (E-Kost) berbasis website.  
3(1), 1–10.  
11. Nugraheny, D., Indrianingsih, Y., Kurniawan, S., & Sunaryo, H. (2023). Prediksi target pendapatan  
minimum CV Bunga Dasimoen Florist menggunakan metode regresi linier berganda berbasis web.  
Angkasa, 15(1), 89–98. https://doi.org/10.28989/angkasa.v15i1.1592  
12. Scikit-learn  
developers.  
(2025).  
LinearRegression  
documentation.  
https://scikit-  
learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html  
13. Setiawan, R., Supriatna, A. D., & Kusuma, A. H. (2021). Rancang bangun sistem informasi  
pengelolaan Rumah Kos Deo Garut berbasis web. Jurnal Algoritma, 17(2), 368–377.  
https://doi.org/10.33364/algoritma/v.17-2.368  
14. Susnjak, T., Ramaswami, G. S., & Mathrani, A. (2022). Learning analytics dashboard: A tool for  
providing actionable insights to learners. International Journal of Educational Technology in  
Higher Education. https://doi.org/10.1186/s41239-021-00313-7  
Page 3951  
www.rsisinternational.org